Net Neutrality, Business Models, and Internet Interconnection

Size: px
Start display at page:

Download "Net Neutrality, Business Models, and Internet Interconnection"

Transcription

1 Net Neutrality, Business Models, and Internet Interconnection Jay Pil Choi Doh-Shin Jeon Byung-Cheol Kim June 6, 2014 Abstract We analyze the effect of net neutrality regulation in a two-sided market framework when content is heterogeneous in its sensitivity to delivery quality. We characterize the equilibrium in a neutral network constrained to offer the same quality vis-à-vis a non-neutral network where Internet service providers (ISPs) are allowed to engage in second degree price discrimination with a menu of quality-price pairs. We find that the merit of net neutrality regulation depends crucially on content providers business models. More generally, our analysis can be considered a contribution to the literature on second-degree price discrimination in two-sided platform markets. JEL Codes: D4, L1, L5 Key Words: Net neutrality, Two-sided markets, Second-degree price discrimination, Content providers business models, Internet interconnection An earlier version of this paper was circulated under the title Internet Interconnection and Network Neutrality (TSE Working Paper, n ). We thank Marc Bourreau, Zhijun Chen, Jane Choi, Philippe Choné, Yuk-Fai Fong, Dominik Grafenhofer, Benjamin E. Hermalin, Bruno Jullien, Teddy Kim, David Laband, Wooyoung Lim, Laurent Linnemer, Qihong Liu, Hyunwoo Park, Yossi Spiegel, Ingo Vogelsang, and participants in various conferences and seminars for helpful comments. We are also grateful to two anonymous referees for constructive comments which greatly improved the paper. Jay Pil Choi s research was supported under Australian Research Council s Discovery Projects funding scheme (project number DP ). Doh-Shin Jeon s research was supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2014S1A5A ). School of Economics, University of New South Wales, Sydney, NSW 2052, Australia, Department of Economics, Michigan State University and School of Economics, Yonsei University. Toulouse School of Economics (GREMAQ, IDEI), Manufacture de Tabacs, 21 Allée de Brienne, 31000, Toulouse, France and School of Economics, Sungkyunkwan University, 53 Myungryun-Dong 3 Ga, Jongro-Gu, Seoul, Korea, School of Economics, Georgia Institute of Technology, 221 Bobby Dodd Way, Atlanta, GA

2 1 Introduction With the emergence of the Internet as a fundamental infrastructure for communication, information and commercial activities, net neutrality has become one of the most important regulatory policy issues. In essence, net neutrality is a principle that Internet service providers (ISPs) must treat all packets equally and deliver them on a first-come, first-served basis without blocking or prioritizing any traffic based on types of Internet content, services, or applications. 1 From its inception, the Internet has implicitly been governed by this principle of equal access to all types of content. However, with the emergence of various online multimedia services that demand a significant amount of network bandwidth, network congestion and efficient management of network resources have become important policy issues. In particular, content and applications differ in their sensitivity to delay in delivery. For instance, data applications such as can be relatively insensitive to moderate delivery delays from users viewpoints. By contrast, streaming video/audio or Voice over Internet Protocol (VoIP) applications can be very sensitive to delay, leading to jittery delivery of content that provides unsatisfactory user experiences. With such heterogeneity concerning delay costs, one may argue that network neutrality treating all packets equally is not an efficient way to utilize the network. In addition, with the reclassification of Internet transmissions from the category of telecommunications services to the category of information services by the Federal Communications Commission (FCC) in 2005, ISPs are no longer considered as a common carrier utility and subject to nondiscrimination restrictions. As a result, major ISPs such as AT&T, Comcast and Verizon have expressed an interest in offering multi-tiered services that would provide content providers paying a premium with a faster access to their networks. In response, the Federal Communications Commission (FCC) implemented three basic Open Internet Rules in 2010 to maintain the status quo of net neutrality regime: transparency, no blocking, and no unreasonable discrimination. 2 However, on January 14, 2014, the U.S. Court of Appeals for the D.C. Circuit vacated the non-discrimination and anti-blocking rules of the FCC s Open Internet Order based 1 There is no universally accepted definition of net neutrality. The definition given here is similar to what Krämer, Wiewiorra and Weinhardt (2013) calls strict net neutrality and consumer groups advocate. For various definitions of net neutrality and more details about the debate on net neutrality, see the survey paper by Krämer et al. (2013). 2 According to the FCC website (http://www.fcc.gov/guides/open-internet), The Open Internet is the Internet as we know it... Network, or net, neutrality is just another way of referring to Open Internet principles. 1

3 on a technical reason that the FCC does not have the authority to regulate how ISPs grant access to content. 3 In this net neutrality debate, consumer groups and policy makers are mostly concerned with the possibility that ISPs may cut deals with some content providers for faster access while forcing others into the slow lane (USA TODAY, January 16, 2014) by offering multi-tiered services, whereas ISPs argue that they should be able to manage their networks as they see fit to justify their investments in upgrading their networks. To address this issue, we analyze the effect of net neutrality regulation in a two-sided market framework in which Internet service providers (ISPs) serve as platforms that connect content providers (CPs) and end consumers. To reflect the current debate, we focus on price discrimination based on the speed of access as a particular deviation from the net neutrality principle. On the CP side, there is a continuum of heterogeneous content/application providers. CPs content differs in its sensitivity to delivery quality: for a clear exposition, we consider two types of CPs. We compare the equilibrium in a neutral network constrained to offer the same quality with the one in a non-neutral network where ISPs are allowed to engage in second degree price discrimination with a menu of qualityprice pairs. While the difference in the congestion sensitivity of content providers justifies the need to provide multiple lanes for different delivery qualities, we find that a neutral treatment can be welfare-enhancing depending on the CPs relative share of total surplus. As the degree of consumer surplus that CPs extract is primarily affected by the business models that CPs use, our results reveal the importance of CPs business models in assessing the effects of net neutrality regulation. More generally, our findings contribute to the literature on second degree price discrimination in two-sided markets. We show that how a platform s second-degree price discrimination fares against no discrimination depends on the relative allocation of each group s surplus in a two-sided market. 4 3 To quote Circuit Judge David Tatel, Given that the commission has chosen to classify broadband providers in a manner that exempts them from treatment as common carriers, the Communications Act expressly prohibits the commission from nonetheless regulating them as such. Because the commission has failed to establish that the anti-discrimination and anti-blocking rules do not impose per se common carrier obligations, we vacate those portions of the Open Internet Order. See Verizon v. FCC & USA, No (D.C. Cir.). As of this writing, the FCC has proposed revised rules that would prevent ISPs from deliberately blocking or slowing content delivery, but allow CPs to pay for a faster lane of service (FCC 14-61, May 15, 2014). Netflix has already cut a deal with Comcast for a faster streaming service. 4 While our model is primarily motivated by the net neutrality debate, its implications may 2

4 To establish our main intuition, we start with a monopolistic ISP facing homogeneous consumers as a basic model. In particular, we first consider a scenario in which the surpluses from interactions between the CPs and end consumers can be entirely appropriated by one-side of the market. In such a scenario, price discrimination is socially more efficient than neutral treatment; thereby we stack the deck against the neutral regime. Nonetheless, we show that the social welfare can be higher with neutrality regulation when surplus extraction is neither full nor zero. The intuition for this main result is as follows. When choosing the quality for low type CPs, the ISP faces a trade-off arising from the two-sided nature of its business. A downward distortion in the quality for low type CPs has a benefit of extracting more rent of high type CPs, on the CP side, and a cost of reducing the consumer surplus that the ISP can extract, on the consumer side. This trade-off implies that as the ISP focuses on extracting consumer surplus rather than CPs surplus, there will be less distortion in quality. Even if the ISP can extract full consumer surplus regardless of the neutrality regulation in place, the ISP tends to focus more on extracting CPs surplus in the non-neutral network than in the neutral network since the ISP has more instruments to extract CPs surplus in a non-neutral network than in a neutral network. This is why welfare can be higher in the neutral network than in the nonneutral network. More specifically, consider first the extreme case in which consumers take the entire surplus generated by content delivery and CPs share is zero. Then, the ISP will provide the first best quality for each type of CPs in a non-neutral network because this allows them to extract the highest consumer surplus from subscription fees. By contrast, a suboptimal single quality is provided in a neutral network due to the regulatory restriction. Thus, in this case the non-neutral network yields a strictly higher social welfare than the neutral network. As the CPs relative share of total surplus increases, extracting high type CPs rent becomes more and more important. As a consequence, the social welfare ranking between the two regimes would be reversed due to the accelerated quality distortion against low type CPs in the non-neutral network, provided that the neutral network still serves both types of CPs. As the CPs relative share further increases, the exclusion of low type CPs may extend to any market environments in which a two-sided platform has a substantial market power in one side of the market and practice menu pricing in the other side. One example is Korean mobile operators which used subsidies to handsets as a price discrimination instrument against mobile phone customers. The major wireless carriers completely control the distribution of mobile phones. The Korea Communications Commission (KCC) decided to regulate on smartphone subsidies to limit price discrimination by capping the amount of subsidies per handset to KRW 270,000. 3

5 occur under a single quality provision and the non-neutral network reclaims a higher social welfare. This is because the non-neutral network still serves low type CPs while high type CPs are offered the first best quality in both network regimes. This shows that the welfare comparison between the two different network regimes may reveal a non-monotonic relationship with respect to the relative allocation of total surplus between CPs and end consumers. Can this result still be meaningful even in the real Internet environment in which multiple ISPs intensely compete and consumers are heterogeneous in their preferences toward different ISPs? Our answer to this inquiry is positive. With competing ISPs and heterogeneous consumers, several additional issues arise. In particular, when both CPs and consumers belong to the same ISP, all traffic can be delivered on-net. However, if a CP purchases a delivery service from one ISP and consumers subscribe to another ISP, interconnection between these two ISPs is required for the completion of content delivery. In addition, even if there is an agreement concerning the desirability of offering multi-tiered Internet services, implementation of such a system is not a simple matter with interconnected networks. Guaranteeing a specified quality (speed) of content delivery requires cooperation from other networks when content providers and end users belong to different networks. We assume that the ISPs agree on the delivery quality and reciprocal access charge(s) for the delivery of other ISPs traffic that terminate on their own networks. We assume that CPs can multi-home whereas consumers single-home and constitute competitive bottlenecks. 5 However, because of the interconnection arrangement, a CP can deliver content to consumers subscribed to different ISPs without subscribing to multiple ISPs. We find that any equilibrium with interconnection is governed by the so-called off-net cost pricing principle on the CP side. The off-net cost pricing principle was discovered by Laffont, Marcus, Rey, and Tirole (hereafter LMRT, 2003) and Jeon, Laffont and Tirole (2004). It means that network operators set prices for their customers as if their customers traffic were entirely off-net. We find a novel equivalence result : competing ISPs agree on access charges and delivery qualities to maximize the same objective as a monopolistic ISP facing homogeneous consumers, even if this implies intensified competition on the consumer side. By using this equivalence, we show that the qualitative results derived with a monopolistic ISP naturally extend to the case of competing ISPs with interconnection. 6 5 See Armstrong (2006) for various modes of competition in two-sided markets. 6 Our analysis of interconnection can be of independent interest and contributes to the interconnection literature by considering heterogeneous content and price discrimination. We generalize 4

6 The remainder of the paper is organized as follows. In Section 2, we discuss related literature and our contribution. In Section 3, we set up a basic model of two-sided markets with a monopolistic ISP and homogeneous consumers. We analyze the effects of price discrimination across different types of content with a menu of contracts. In Section 4, we compare a neutral network with a non-neutral network in terms of quality choices and social welfare and derive conditions under which the neutral regime can provide higher welfare than the non-neutral one. Section 5 extends the analysis to competing interconnected ISPs facing heterogeneous consumers. We derive a central equivalence result between competing ISPs and a monopolistic ISP. This equivalence result ensures the robustness of our result to the introduction of ISP competition and interconnection. Section 6 contains our concluding remarks. 2 Related Literature Our research contributes to the literature on net neutrality. With net neutrality being one of the most important global regulatory issues concerning the Internet, there has been a steady stream of academic papers on various issues associated with net neutrality regulation in recent years. 7 Hermalin and Katz (2007) examine a situation in which ISPs serve as platforms to connect CPs with consumers. Without any restrictions, ISPs can potentially offer a continuum of vertically differentiated services to a continuum of types of CPs. With net neutrality regulation, ISPs are required to provide a single tier of Internet service. They compare the single service equilibrium with the multi-service equilibrium. One novelty of our paper with respect to Hermalin and Katz (2007) is that we analyze how the relative merit of allowing second-degree price discrimination depends on CPs business models that determine the relative allocation of total surplus between CPs and end consumers. In addition, we extend the analysis to allow for interconnection between competing ISPs. 8 To the best of our knowledge, we are the first to explore implications of net neutrality in the framework of two-sided markets with interconnected and competing the finding of LMRT to a setting of heterogeneous content with different delivery qualities across content. 7 See Lee and Wu (2009), Schuett (2010), Krämer et al. (2013) for the surveys about economics literature on network neutrality. 8 Although they model a two-sided market, they would obtain the same qualitative results in a one-sided market. In contrast, in our model, a non-neutral network always generates higher welfare than a neutral network if the market is one-sided while this can be reversed if the market is two-sided. 5

7 ISPs. Choi and Kim (2010) analyze the effects of net neutrality regulation on investment incentives of a monopoly ISP and CPs. They show that ISPs may invest less in capacity in a non-neutral network than in a neutral network because expanding capacity reduces the CPs willingness to pay for having a prioritized service. Economides and Hermalin (2012) derive conditions under which network neutrality would be welfare superior to any feasible scheme for prioritized service given a capacity of bandwidth. They show that the ability to price discriminate enhances incentives to invest, creating a trade-off between static and dynamic efficiencies. As these papers consider a monopolistic ISP, the interconnection and competition issues do not arise. Bourreau, Kourandi, and Valletti (2012) analyze the effect of net neutrality regulation on capacity investments and innovation in the content market with competing ISPs. They show that investments in broadband capacity and content innovation are higher under a non-neutral regime. However, they do not allow interconnection between ISPs and assume that a CP has access only to end users connected to the same ISP. Economides and Tåg (2012) also consider both a monopolistic ISP and duopolistic ISPs. Once again the issue of Internet interconnection is not considered as they focus on how net neutrality regulation as a zero pricing rule affects pricing schemes on both sides of the market and social welfare. Our research also relates to LMRT (2003) who analyze how the access charge allocates communication costs between CPs and end consumers and thus affects competitive strategies of rival networks in an environment of interconnected networks. They show that the principle of off-net cost pricing prevails in a broad set of environments. Our model builds upon their interconnection model, but focuses on the provision of optimal quality in content delivery services by introducing heterogeneity in CPs content types. In this setting, we analyze how the quality levels and access charges are determined (depending on CPs business models and on whether there exists net neutrality regulation) and find a novel equivalence result between competing ISPs and a monopolistic ISP. There is a large literature on interconnection in the telecommunication market, initiated by Armstrong (1998) and Laffont, Rey, and Tirole (1998a,b). These researchers show that if firms compete in linear prices, they agree to set interconnection charges above associated costs to obtain the joint profit-maximizing outcome and derive the welfare-maximizing interconnection charge that is lower than the privately negotiated level. They also show that the nature of competition can be altered significantly depending on whether or not two-part tariffs or termination-based price discrimination are employed as price instruments. Their models, however, are devoid of the 6

8 issue of transmission of quality because all calls are homogeneous. In contrast, we consider heterogeneous types of CPs requiring different transmission qualities and analyze quality distortions associated with net neutrality regulations. Armstrong (1998) and Laffont, Rey, and Tirole (1998a,b) consider inelastic subscription of consumers while we consider elastic subscription demand. If we consider inelastic subscription, we find a profit neutrality result. 9 Although some papers study elastic subscription demand in the literature on interconnection in the telecommunications market (see Armstrong and Wright, 2009, Dessein, 2003, Hurkens and Jeon, 2012), to our knowledge, our result of equivalence between competing ISPs and a monopolistic ISP is new. 3 A Monopolistic ISP in a Two-sided Market 3.1 ISPs, CPs, and Consumers We consider a model of a two-sided market to analyze the effects of price discrimination on various market participants and social welfare. To be concrete, we consider a monopolistic ISP that serves as a platform in a two-sided market where CPs and end consumers constitute two distinct groups of customers. As pointed out by LMRT (2001, 2003), the traffic between CPs and the traffic between consumers take up small volumes relative to the volume of traffic from CPs to consumers. Thus, we focus on the primary traffic from CPs to consumers who browse web pages, download files, stream multi-media content, etc. 10 There is a continuum of CPs whose mass is normalized to one. We consider a simple case of CP heterogeneity. There are two types of CPs: θ {H, L}, with = H L > 0. The measure of θ type CP is denoted by ν θ, where ν H = ν and ν L = 1 ν. There is also a continuum of consumers who demand one unit of each content whose value depends on content type θ and its quality q. In our 9 More precisely, in an earlier version of this paper where we assume that consumers are distributed over a Hotelling line, we showed that each ISP s profit is constant and equal to the Hotelling profit regardless of quality levels and access charges. 10 For instance, exchanges between consumers take up trivial volumes. With the developments of VoIP, file sharing, and online gaming services, the absolute volume of consumerto-consumer traffic is increasing, but still expected to constitute a relative decrease of the percentage of total traffic due to explosive growth in Internet video streaming and downloads. Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, , available at paper c pdf See 7

9 context, quality means speed and reliability of content delivery. Let q θ denote the quality of delivery associated with content of type θ. The total surplus generated from interaction between a consumer and a CP of type θ is equal to θu(q), where u > 0 and u < 0 with the Inada condition lim q 0 u (q) =. According to our utility formulation, θ reflects the sensitivity of content to delay, with higher valuation content being more time/congestion sensitive. 11 Note that θu(q) captures not only a consumer s gross surplus but also a CP s revenue from advertising. We assume that this surplus is divided between a CP and a consumer such that the former gets αθu(q) and the latter (1 α)θu(q) with α [0, 1]. The parameter α reflects the nature of the CPs business model. We have in mind two sources of revenue for CPs: micropayments and advertising revenue. 12 For instance, the parameter α would be higher if CPs can extract surplus from consumers via micropayments in addition to advertising revenues. If the CPs revenue source is limited to advertising, α can be relatively low. We later show that the CPs business model, captured by α, plays an important role in assessing the effects of net neutrality regulations. A monopolistic ISP provides content delivery service from CPs to consumers. The marginal cost of providing a unit traffic of quality q from CP to end users is assumed to be linear, i.e., c(q) = cq for q We consider two different regimes under 11 For our analysis, the type space is essentially about the sensitivity to transmission quality; perfect correlation between the value of content and its sensitivity to transmission quality is not necessary. To illustrate this, consider an alternative formulation that explicitly accounts for the transmission delay costs. For instance, assume that all CPs provide the same gross utility of v (i.e., there is only one type in terms of the content value), but this value is reduced with a delay. Consider an additive specification in which the total surplus generated from interaction between a consumer and a CP of type θ is equal to v θd, where d is delay time. With this additive specification (and independence between the content value and the sensitivity type parameter), we can derive exactly the same qualitative results as in the paper. 12 Our simplifying assumption is that CPs are homogeneous in all dimensions except for their type θ. This implies that they use the same business model; otherwise, CPs have two-dimensional types (θ and business model) from the ISP s point of view. In an ad-based business model, we can assume that the advertising revenue is proportional to consumer gross utility, which in turn is proportional to θu(q). Then, the total surplus is given by bθu(q) where b is a positive constant. Hence, by redefining bθ as θ, we are back to our original formulation. See Appendix B for a micro-foundation of such an advertising model. 13 The assumption of a linear marginal cost in quality can be made without any loss of generality because we can normalize quality to satisfy the assumption of linearity. Suppose that c(q) is nonlinear. By redefining q as c(q)/c, we have a linear marginal cost function c( q)= c q. Starting from a concave utility function and a convex cost function, the utility function with the normalized cost still remains concave. 8

10 which ISPs can deliver content: a neutral regime or a non-neutral regime. Under a non-neutral regime, the ISP can offer multiple classes of services that differ in delivery quality. We assume that the ISP is unable to practice first-degree price discrimination across content providers depending on content types, but can engage in second degree price discrimination by offering a menu of contracts that charges different prices depending on the quality of delivery. Let q H be the quality for hightype CPs and q L for low-type CPs. In a neutral regime or in the presence of net neutrality regulation, ISP i is constrained to offer a single uniform delivery quality q. To focus on the effects of the ISP s price discrimination against CPs, we assume homogeneous consumers. This setup allows the ISP to extract the whole consumer surplus. However, we extend the analysis later to allow for competing interconnected ISPs facing heterogeneous consumers with elastic participation, and establish an equivalence result between the monopolistic outcome of the basic model and the outcome of competing ISPs in terms of offers made to the content side. Let U(α) denote the gross utility a consumer derives from the content side by subscribing to the ISP. We have U(α) = u + (1 α) ν θ θu(q θ ), where u is the θ intrinsic utility associated with the Internet connection. A consumer s net utility from subscribing to the ISP, u, is given by u(α, f) = U(α) f, (1) where f is the subscription price charged by the ISP. We normalize, without loss of generality, the total measure of consumers to one. Finally, we assume that the monopoly ISP simultaneously announces the price-quality pairs for CPs and the fee for consumers. 14 Before analyzing market outcomes under various regimes, we first analyze the firstbest outcome as a benchmark. It is clear that the socially optimal quality should maximize θu(q) cq and hence, the first-best quality level for CPs of type θ, denoted qθ F B, is determined by the following condition: θu (q F B θ ) = c. (2) 14 The optimal outcome chosen by the monopoly ISP in this simultaneous pricing is the same as the one chosen in a sequential pricing in which it first chooses the price-quality pairs for CPs and then the fee for consumers. 9

11 The marginal benefit of an incremental improvement of delivery quality for the content of type θ must be equal to c, the marginal cost associated with such an adjustment. With heterogeneous content that differs in sensitivity to delivery quality, the uniform treatment of content mandated by net neutrality in general would not yield a socially optimal outcome. Our modeling strategy is driven by institutional features of the Internet market. First, in many countries including the US, customers predominantly pay flat fees for their access to the Internet regardless of the traffic amount they generate. 15 In addition, even if consumers pay different prices depending on the amount of data they stream, differential pricing on the consumer side seems to be non-controversial and even considered necessary to manage over-loaded Internet traffic. Thus, we focus on two-tiered services with price discrimination in only the content provider side Non-neutral Network and Second-degree Price Discrimination Let {(p H (α), q H (α)), (p L (α), q L (α))} be the menu of contracts offered to CPs which satisfies the incentive and participation constraints of CPs (defined below). Then, each consumer s gross utility is given by U(α) = u + (1 α) ν θ θu(q θ ), which can θ be fully extracted by a subscription fee f as consumers are homogeneous. The ISP s profit from the content side is π CP = ν θ [p θ cq θ ]. The overall profit for the ISP θ can be written as Π M (α) = U(α) + π CP. Thus, the monopolistic ISP s mechanism design problem can be described as: max Π M (α) = u + (p θ,q θ ) θ ν θ [p θ + (1 α)θu(q θ ) cq θ ] subject to IC H : αhu(q H ) p H αhu(q L ) p L ; IC L : αlu(q L ) p L αlu(q H ) p H ; IR H : αhu(q H ) p H 0; IR L : αlu(q L ) p L 0, 15 See Krämer et al. (2013) for the flat rate trap the ISPs are in on the consumer side of the market. 16 However, we verify the robustness of our results in a simple extension of the model with two symmetric sides in which both the end user and the CP sides are heterogeneous and the ISP practices menu pricing on both sides (the details are available upon request). 10

12 where IC θ and IR θ refer to type θ CPs incentive compatibility constraint and individual rationality constraint, respectively. This is a standard mechanism design problem for second-degree price discrimination. As usual, the high-type s incentive compatibility constraint IC H and the low-type s individual rationality constraint IR L are binding: we thus have p H = αhu(q H ) α u(q L ); p L = αlu(q L ). (3) This leads to the following reduced problem max {q H,q L } Π M (α, q H (α), q L (α)) = u + θ ν θ [θu(q θ ) cq θ ] αν u(q L ). The objective in the reduced program shows that the ISP extracts full surplus except for the rent to high type CPs, which is given by αν u(q L ). Let {(p H (α), q H (α)), (p L (α), q L (α))} be the menu of contracts chosen by the ISP under a non-neutral network. From the first order conditions, we find that the optimal quality for the high type is determined by Hu (qh ) = c for any α, which is equal to the first-best level, regardless of α, i.e., qh = qf H B. By contrast, the low type CPs quality is characterized by ( L ν ) 1 ν α u (ql(α)) = c. (4) As in the standard mechanism design problem, there is a downward distortion in quality for the low type, that is, ql B (α) qf L α = 0. with the equality holding only for We assume that if CPs extract all the surplus from consumers (i.e., α = 1), the monopoly ISP prefers serving both types under second-degree price discrimination. Assumption 1. ql (α = 1) > 0 Under the Inada condition, ql (α = 1) > 0 is equivalent to L > ν. Assumption 1 1 ν ensures that ql (α) > 0 for any α [0, 1] because total differentiation applied to (4) shows that the low-type quality is decreasing in α : dql dα = ν u (ql ) < 0. (5) ((1 ν)l αν ) u (ql ) For a given q L, the rent obtained by a high type CP increases with α. Hence, as the CPs share of surplus (i.e., α) increases, the ISP has more incentives to distort the quality for the low type. From the envelope theorem, the maximized objective under 11

13 a non-neutral network strictly decreases with α: dπ M (α, q H (α), q L (α)) dα = ν u(q L ) < Neutral Network and No Price Discrimination Now consider a neutral network where the ISP is constrained to choose only a single price-quality pair (p, q). Given this single quality offer constraint, the ISP decides between serving only the high type CPs with the exclusion of the low type CPs and serving both types of CPs. With the exclusion, it is straightforward that the ISP will choose q = qh F B B and p = αhu(qf H ), which gives Π EX = u + ν[hu(qh F B) cqf H B].17 If the monopolistic ISP decides to serve both types, then p = αlu(q) and f = u + (1 α) θ ν θθu(q). Hence, the monopoly ISP chooses a single quality q to solve max q Π(α, q) = u + (L + (1 α)ν )u(q) cq. From the first-order condition, we obtain the optimal quality choice when both types of CPs are served: (L + (1 α)ν )u ( q(α)) = c. (6) Equation (6) indicates that the optimal quality choice lies between the first-best level qualities for the high and the low types, that is, q F B L q(α) < qh F B. If the ISP could. However, we extract CPs surplus only, then q(α) could never be higher than q F B L find that q(α) is strictly higher than ql F B for any α < 1. This has to do with the two-sided nature of the ISP s business: since it can extract extra consumer surplus generated by high type CPs in addition to Lu(q), it chooses a quality level higher than ql F B. By total differentiation of (6), we can derive that the quality decreases with α: d q(α) dα = ν u ( q) < 0. (7) (L + (1 α)ν )u ( q) Let Π(α) Π(α, q(α)). From the envelope theorem, the monopolistic ISP s profit without exclusion strictly decreases with α as in the non-neutral regime. d Π(α) dα = ν u( q(α)) < 0 By contrast, the ISP s profit under exclusion, Π EX, is independent of α. We assume 17 We use a tilde ( ) to denote variables associated with a neutral network. 12

14 the following. Assumption 2. Π(α = 0) > Π EX > Π(α = 1) This assumption is made to highlight the difference between one-sided markets and two-sided markets in evaluating the relative merit of no price discrimination (PD) vis-à-vis second-degree PD: even if second-degree PD generates higher welfare than no PD in one-sided markets, it does not necessarily mean that the same holds for two-sided markets. Assumption 2 is introduced to make this point clear. 18 addition, Assumption 2 is introduced to reduce the number of cases to consider. relevant, we will comment on how the details of the analysis can change when this assumption is violated. Assumption 2, together with the monotonicity of Π(α), implies that there exists a unique threshold level of α denoted by α N (0, 1) such that the monopolistic ISP serves both types of CPs for α < α N and excludes the low type CPs for α > α N, where α N is implicitly defined by Π(α N ) = Π EX, that is, (L + (1 α N )ν )u( q(α N )) c q(α N ) = ν (Hu(qH F B ) cqh F B ). (8) Therefore, the monopolist ISP s profit under the neutral system, ΠM (α), can be written as Π M (α) = and can be shown as in Figure 1. { Π(α) for α < α N Π EX for α α N Let ( p (α), q (α)) represent the ISP s choice under a neutral network. Then, the quality chosen by the ISP is given by: q (α) = { q(α) for α < α N q F B H for α αn The corresponding retail prices are given by p (α) = αlu( q(α)) for α < α N p (α) = αhu(q F B H ) for α > αn Note that Assumption 2 is not necessary if our purpose is only to show the possibility that net neutrality may be superior to non-neutrality. If Assumption 2 is violated, Π(α = 1) > Π EX (i.e., no exclusion occurs under a neutral network for α = 1). In such a case, a sufficient condition for the result is that a neutral network generates a strictly higher welfare than a non-neutral network for α = 1. Then, by the continuity argument, the neutral network generates a higher welfare than the non-neutral network for α close to If Assumption 2 is violated and Π(α = 1) > Π EX, there will be no exclusion of CPs. We thus have Π M (α) = Π(α), q (α) = q(α), and p (α) = αlu( q(α)) for all α. In If and 13

15 Figure 1: The monopolist ISP s profit in the neutral network 3.4 Assumptions and Social Welfare with One-sided Market Under Assumptions 1-2, the non-neutral network dominates the neutral network, from the social welfare point of view, for the extreme cases of α = 1 and α = 0. Essentially, these two cases can be considered as representations of one-sided markets. Consider first the case in which CPs capture the whole surplus from interactions with consumers, i.e., α = 1. Then, each consumer obtains the basic utility u only. So, the monopoly ISP will set f = u both under non-neutral and neutral networks. Consequently, we can focus on the monopoly ISP s problem of maximizing profit from CPs, which is a standard problem in one-sided markets. In this case, high type CPs consume qh F B in both regimes, but low types are served only under a non-neutral network. This is a standard argument in favor of second-degree price discrimination. For the other extreme case of α = 0, consumers capture all surplus from interactions with CPs. Since consumers are homogeneous, the monopoly ISP can extract full surplus from consumers. The case of α = 0 is the same as a standard monopoly in a one-sided market with cost function cq. The monopoly will provide services for free to CPs, which means that the ISP bears the entire cost of cq. Define u F B and c F B as the gross utility from all content providers and its associated content delivery cost for each consumer when the first best delivery qualities are chosen: u F B = θ c F B = θ ν θ θu(q F B θ ), (9) ν θ cq F B θ. (10) 14

16 Under a non-neutral regime, the monopoly ISP provides the first-best quality for each type of CPs and charges the consumer subscription fee f(α = 0) = u + u F B. 20 By contrast, the monopoly ISP is constrained to offer one level of quality under a neutral regime and hence can never achieve the first-best outcome. In summary, we have: Proposition 1. Consider a monopoly ISP facing homogeneous consumers with inelastic subscription. (i) If α = 1, under Assumptions 1-2, the ISP serves both types of CPs in a nonneutral network but serves only high types in a neutral network. Therefore, social welfare is higher under a non-neutral network than under a neutral network. (ii) If α = 0, the outcome chosen by the ISP coincides with the first-best under a non-neutral network. By contrast, under a neutral network, the first-best can never be realized. Therefore, social welfare is higher under a non-neutral network than under a neutral network. Under Assumptions 1-2, we consider a scenario in which the neutral network is always dominated in one-sided market settings, stacking the deck against the neutral network. This result will be contrasted to the case where a neutral network can provide a higher social welfare relative to a non-neutral network, as we consider intermediate values of α. The parameter α represents the surplus division between CPs and end users when they interact through the ISP and indicates which side the ISP should focus on to extract rents. As α increases, CPs capture more surplus and the extraction of rents from the CP side becomes more important. As a result, the ISP distorts the quality for low type CPs further down to reduce the rent of the high type CPs under the nonneutral network. Under the neutral network, ISPs exclude the low type CPs when α is high enough (i.e., α > α N ) to reduce the rents of high type CPs under Assumption 2. This finding may have important policy implications. For instance, α may capture how much CPs can extract consumer surplus through micropayments. From this perspective, the concern about potential exclusion of CPs in a neutral network can be heightened if the business model of CPs shifts from an advertising-based one with free access to the one with micropayments that directly charges consumers for content. 20 When α = 0, every CP makes zero profit and we can assume that a CP follows the ISP s desire in the case of indifference. For any α > 0 (hence α can be as close as possible to zero) and q > 0, the ISP can exclude low types by charging p = αhu(q). 15

17 So far, we have treated α as a parameter. In Appendix B, we provide a very simple informative advertising model as a micro-foundation to illustrate how α can be endogenized and affected by the development of targeted advertising. For instance, it is shown that if targeted advertising induces more effective advertising but solicits more privacy concerns due to its intrusiveness, α increases. 4 Comparison of Neutral vs. Non-neutral Networks In this section, we compare quality choices and social welfare in the neutral network to those in the non-neutral network. 4.1 Quality Choices and Each Group s Payoff Figure 2 shows the optimal quality schedules for both network regimes. In a nonneutral network, there is no distortion in the quality for high type CPs and a downward distortion in the quality for low type CPs. As α decreases, this distortion becomes smaller and becomes zero when α = 0 (i.e., q (α = 0) = ql F B ). In a neutral network, the ISPs serve only high type CPs for α > α N and choose q (α) = q F B; for α α N, they choose a quality q (α) ( ql F B decreases with α., qf B H ) and serve both types where q (α) The ISP realizes a strictly higher profit in a non-neutral network than in a neutral network by the revealed preference argument; in a non-neutral network the ISPs could always choose an equal quality for both delivery services if this would give a higher profit. Consumers are indifferent across the two regimes because their surplus is completely extracted by the ISP anyway. 21 Low type CPs always receive zero rents for any α regardless of net neutrality regulation. Comparison of high type CPs payoff depends on whether α is higher or lower than α N. If α α N, the relationship of q (α) > ql F B ql (α) implies that high type CPs obtain a higher payoff in the neutral network than in the non-neutral network. If α > α N, the reverse holds since high type CPs obtain no rent in the neutral network while they obtain a strictly positive rent in the non-neutral network. 21 However, this result is due to the assumption of homogeneous consumers. In an extension with competing ISPs and heterogeneous consumers, we show that consumer surplus is higher in the non-neutral network than in the neutral network. H 16

18 Figure 2: The optimal quality schedules 4.2 Social Welfare We now perform a welfare comparison to assess the merit of net neutrality regulations that prohibit price discrimination in two-sided markets. Recall that under Assumptions 1 and 2, a neutral network cannot outperform a non-neutral network for the two extreme cases of full or zero extraction of surplus by CPs vis-à-vis consumers (see Proposition 1). We investigate whether this result is robust to intermediate cases of α (0, 1) and find that the social welfare ranking between the two regimes can be reversed for intermediate values of α. As we provided the intuition for this result in the introduction, the single quality restriction can be welfare-enhancing because of its smaller quality distortion in a neutral network compared to a non-neutral network, despite offering a suboptimal quality for the high type CPs. Note that the ISP has two sources of revenues: the one from the CP side and the one from the consumer side. When the ISP chooses the quality for low type CPs, it faces a trade-off which arises from the two-sided nature of the ISP s business. More precisely, a downward distortion in the quality for low type CPs has the benefit of extracting more rent from high type CPs, on the content side, and the cost of extracting less consumer surplus, on the consumer side. This implies that when the ISP focuses on making revenue from the consumer side rather than from the CP side, there will be less distortion in quality. Because the ISP has more instruments to extract CPs surplus in a non-neutral network than in a neutral network, it focuses relatively more on extracting CPs surplus in a non- 17

19 neutral network than in a neutral network. In our model, this arises especially when α becomes smaller than α N. For α ( 0, α N), the ISP always downwardly distorts the quality for low type CPs in a non-neutral network whereas in a neutral network, it serves both types of CPs with a quality superior to ql F B. For this reason, a neutral network may provide a higher social welfare than a non-neutral network. 22 Given this insight, we now provide more rigorous mathematical derivation for our result. The social welfare in the non-neutral network with optimal quality choices can be written as W = u+ θ ν θ [θu(q θ) cq θ] = u+ν(hu(q F B H ) cqh F B )+(1 ν)(lu(ql(α)) cq L(α)). We take the first-order derivative of the social welfare with respect to α as (11) dw dα = d[(1 ν)(lu(q L (α)) cq L (α))]. (12) dα Using the first-order optimal quality condition for the low type CPs in (4), we can rewrite (12) as follows: dw dα = αν u (q L) dq L dα < 0. (13) The inequality above holds because the quality distortion increases in α, i.e., dq L dα < 0. Similarly, we can define social welfare under the neutral network and, for the same reason as in the non-neutral network, we find that the social welfare in the neutral network also decreases in α for any α < α N : d W dα = αν u ( q ) d q dα < 0. (14) Recall that the quality adjustment to the change in α can be derived as (5) for the non-neutral network and (7) for the neutral one. To facilitate the comparison further and gain more intuition, let us consider a utility function with Arrow-Pratt constant absolute risk aversion (CARA), e.g., u(q) = A B exp( rq) where r measures the degree of risk aversion with positive r 22 If we relax Assumption 2 and consider the case where a neutral network entails no exclusion and the quality distortion effect is high enough in a non-neutral network, social welfare may be higher in a neutral network when α is equal to 1 (and close to 1, by continuity). 18

20 constants A and B. Then, we can obtain a clear comparison of dql (α) dα > d q dα for ν (0, 1) (15) from u ( q ) = u (q u ( q L ) = r. This implies that the ISP s quality degradation gradient ) u (ql ) for low types in the non-neutral network is steeper than the one for the uniform quality in the neutral network as α increases. In addition, we find u (q L ) > u ( q ) from q L < q for any utility function with u < 0. Hence, we find that the social welfare decreases more quickly as α increases in the non-neutral network compared to the neutral network, i.e., dw dα > d W dα. Given this understanding, let us finally compare the level of social welfare under two different network regimes. Recalling the definition of α N in (8), social welfare in the neutral network at α = α N can be expressed as W α=α N = u + (L + (1 α N )ν )u( q ) c q + α N ν u( q ) = u + ν (Hu(qH F B ) cqh F B ) + α N ν u( q ) This simplifies the comparison between W and W evaluated at α = α N comparison between α N ν u( q ) and (1 ν)(lu(q L ) cq L ) : as the W α=α N > W α=α N α N ν u( q ) > (1 ν)(lu(q L) cq L) (16) Under Assumptions 1-2, social welfare can be higher in a neutral network relative to a non-neutral network as long as (16) is satisfied. Since W α=0 < W α=0 and dw dα > d W dα, if (16) is satisfied, there exists a unique α such that W > W for any α (α, α N ]. Proposition 2. Consider the CARA utility function for which Assumptions 1 and 2, and (16) are satisfied. There exists a unique level of α, which is weakly smaller than α N but is greater than zero, such that social welfare is higher in the neutral network than in the non-neutral network for all α (α, α N ]. Figure 3 illustrates a plausible case in which a neutral network may yield a higher total social welfare than a non-neutral network. As a numerical exercise, consider parameters such that H = 40, L = 30, u = 5, and ν = If we consider a CARA utility function such as u(q) = e 2q, it satisfies all assumptions that we make and the neutral network yields the higher social welfare for α (0.2441, ) than 19

Net Neutrality, Business Models, and Internet Interconnection

Net Neutrality, Business Models, and Internet Interconnection Net Neutrality, Business Models, and Internet Interconnection Jay Pil Choi Doh-Shin Jeon Byung-Cheol Kim July 18, 2013 Abstract We analyze the effect of net neutrality regulation in a two-sided market

More information

Net Neutrality, Business Models, and Internet Interconnection

Net Neutrality, Business Models, and Internet Interconnection Net Neutrality, Business Models, and Internet Interconnection Jay Pil Choi Doh-Shin Jeon Byung-Cheol Kim March 24, 2014 Abstract We analyze the effect of net neutrality regulation in a two-sided market

More information

Internet Interconnection and Network Neutrality

Internet Interconnection and Network Neutrality 12-355 November 2012 Internet Interconnection and Network Neutrality Jay Pil Choi, Doh-Shin Jeon and Byung-Cheol Kim Internet Interconnection and Network Neutrality Jay Pil Choi Doh-Shin Jeon Byung-Cheol

More information

Internet Interconnection and Network Neutrality

Internet Interconnection and Network Neutrality Internet Interconnection and Network Neutrality Jay Pil Choi Doh-Shin Jeon Byung-Cheol Kim June 19, 2012 Abstract This paper analyzes competition between interconnected networks when content is heterogeneous

More information

Strategic Investment Dependence and Net Neutrality

Strategic Investment Dependence and Net Neutrality Strategic Investment Dependence and Net Neutrality by Martin Nielsen Discussion Papers on Business and Economics No. 11/2015 FURTHER INFORMATION Department of Business and Economics Faculty of Business

More information

Lecture 6: Price discrimination II (Nonlinear Pricing)

Lecture 6: Price discrimination II (Nonlinear Pricing) Lecture 6: Price discrimination II (Nonlinear Pricing) EC 105. Industrial Organization. Fall 2011 Matt Shum HSS, California Institute of Technology November 14, 2012 EC 105. Industrial Organization. Fall

More information

Two Papers on Internet Connectivity and Quality. Abstract

Two Papers on Internet Connectivity and Quality. Abstract Two Papers on Internet Connectivity and Quality ROBERTO ROSON Dipartimento di Scienze Economiche, Università Ca Foscari di Venezia, Venice, Italy. Abstract I review two papers, addressing the issue of

More information

Economic Features of the Internet and Network Neutrality

Economic Features of the Internet and Network Neutrality Forthcoming, The Oxford Handbook on the Economics of Networks Economic Features of the Internet and Network Neutrality By Nicholas Economides* April 2015 Abstract We discuss the issue of a possible abolition

More information

Market Power and Efficiency in Card Payment Systems: A Comment on Rochet and Tirole

Market Power and Efficiency in Card Payment Systems: A Comment on Rochet and Tirole Market Power and Efficiency in Card Payment Systems: A Comment on Rochet and Tirole Luís M. B. Cabral New York University and CEPR November 2005 1 Introduction Beginning with their seminal 2002 paper,

More information

Network Neutrality and Network Management Regulation: Quality of Service, Price Discrimination, and Exclusive Contracts

Network Neutrality and Network Management Regulation: Quality of Service, Price Discrimination, and Exclusive Contracts IFN Working Paper No. 890, 2011 Network Neutrality and Network Management Regulation: Quality of Service, Price Discrimination, and Exclusive Contracts Nicholas Economides and Joacim Tåg Research Institute

More information

Foreclosure, Entry, and Competition in Platform Markets with Cloud

Foreclosure, Entry, and Competition in Platform Markets with Cloud Foreclosure, Entry, and Competition in Platform Markets with Cloud Mark J. Tremblay Department of Economics Michigan State University E-mail: trembl22@msu.edu August 27, 2015 Abstract Platforms in two-sided

More information

Hyun-soo JI and Ichiroh DAITOH Tohoku University. May 25, 2003. Abstract

Hyun-soo JI and Ichiroh DAITOH Tohoku University. May 25, 2003. Abstract Interconnection Agreement between Internet Service Providers and the Optimal Policy Intervention: The Case of Cournot-type Competition under Network Externalities Hyun-soo JI and Ichiroh DAITOH Tohoku

More information

Second degree price discrimination

Second degree price discrimination Bergals School of Economics Fall 1997/8 Tel Aviv University Second degree price discrimination Yossi Spiegel 1. Introduction Second degree price discrimination refers to cases where a firm does not have

More information

Network Neutrality and Congestion Sensitive Content Providers: Implications for Content Variety, Broadband Investment and Regulation

Network Neutrality and Congestion Sensitive Content Providers: Implications for Content Variety, Broadband Investment and Regulation Forthcoming at Information Systems Research manuscript ISR-2011-069 Network Neutrality and Congestion Sensitive Content Providers: Implications for Content Variety, Broadband Investment and Regulation

More information

A Strategic Guide on Two-Sided Markets Applied to the ISP Market

A Strategic Guide on Two-Sided Markets Applied to the ISP Market A Strategic Guide on Two-Sided Markets Applied to the ISP Market Thomas CORTADE LASER-CREDEN, University of Montpellier Abstract: This paper looks at a new body of literature that deals with two-sided

More information

Price Discrimination: Part 2. Sotiris Georganas

Price Discrimination: Part 2. Sotiris Georganas Price Discrimination: Part 2 Sotiris Georganas 1 More pricing techniques We will look at some further pricing techniques... 1. Non-linear pricing (2nd degree price discrimination) 2. Bundling 2 Non-linear

More information

Why do merchants accept payment cards?

Why do merchants accept payment cards? Why do merchants accept payment cards? Julian Wright National University of Singapore Abstract This note explains why merchants accept expensive payment cards when merchants are Cournot competitors. The

More information

Prices versus Exams as Strategic Instruments for Competing Universities

Prices versus Exams as Strategic Instruments for Competing Universities Prices versus Exams as Strategic Instruments for Competing Universities Elena Del Rey and Laura Romero October 004 Abstract In this paper we investigate the optimal choice of prices and/or exams by universities

More information

Moral Hazard. Itay Goldstein. Wharton School, University of Pennsylvania

Moral Hazard. Itay Goldstein. Wharton School, University of Pennsylvania Moral Hazard Itay Goldstein Wharton School, University of Pennsylvania 1 Principal-Agent Problem Basic problem in corporate finance: separation of ownership and control: o The owners of the firm are typically

More information

Unraveling versus Unraveling: A Memo on Competitive Equilibriums and Trade in Insurance Markets

Unraveling versus Unraveling: A Memo on Competitive Equilibriums and Trade in Insurance Markets Unraveling versus Unraveling: A Memo on Competitive Equilibriums and Trade in Insurance Markets Nathaniel Hendren January, 2014 Abstract Both Akerlof (1970) and Rothschild and Stiglitz (1976) show that

More information

When to Use the Open Business Model for Software Products under Network Effects?

When to Use the Open Business Model for Software Products under Network Effects? When to Use the Open Business Model for Software Products under Network Effects? Lizhen Xu, Marius F. Niculescu and D. J. Wu Scheller College of Business, Georgia Institute of Technology, Atlanta, GA 30308

More information

PRICE DISCRIMINATION Industrial Organization B

PRICE DISCRIMINATION Industrial Organization B PRICE DISCRIMINATION Industrial Organization B THIBAUD VERGÉ Autorité de la Concurrence and CREST-LEI Master of Science in Economics - HEC Lausanne (2009-2010) THIBAUD VERGÉ (AdlC, CREST-LEI) Price Discrimination

More information

Part IV. Pricing strategies and market segmentation

Part IV. Pricing strategies and market segmentation Part IV. Pricing strategies and market segmentation Chapter 9. Menu pricing Slides Industrial Organization: Markets and Strategies Paul Belleflamme and Martin Peitz Cambridge University Press 2010 Chapter

More information

Platform Competition under Asymmetric Information

Platform Competition under Asymmetric Information Platform Competition under Asymmetric Information Hanna Hałaburda Yaron Yehezkel Working Paper 11-080 Copyright 2011 by Hanna Hałaburda and Yaron Yehezkel Working papers are in draft form. This working

More information

DOC NO: INFOSOC 52/14 DATE ISSUED: June 2014. Resolution on the open and neutral Internet

DOC NO: INFOSOC 52/14 DATE ISSUED: June 2014. Resolution on the open and neutral Internet DOC NO: INFOSOC 52/14 DATE ISSUED: June 2014 Resolution on the open and neutral Internet Introduction This resolution builds on the TACD net neutrality resolution of April 2010 1, which called for policies

More information

e-book Platform Competition in the Presence of Two-Sided Network Externalities

e-book Platform Competition in the Presence of Two-Sided Network Externalities 2012 45th Hawaii International Conference on System Sciences e-book Platform Competition in the Presence of Two-Sided Network Externalities Yabing Jiang Graduate School of Business Administration Fordham

More information

video case e-commerce. business. technology. society. KENNETH C. LAUDON AND CAROL G. TRAVER Issues in E-commerce for You

video case e-commerce. business. technology. society. KENNETH C. LAUDON AND CAROL G. TRAVER Issues in E-commerce for You e-commerce. business. technology. society. KENNETH C. LAUDON AND CAROL G. TRAVER video case chapter 8 Ethical, Social, and Political Issues in E-commerce case 2 What Net Neutrality Means for You watch

More information

Network Neutrality, Access to Content and Online Advertising. RSCAS 2013/57 Robert Schuman Centre for Advanced Studies

Network Neutrality, Access to Content and Online Advertising. RSCAS 2013/57 Robert Schuman Centre for Advanced Studies RSCAS 013/57 Robert Schuman Centre for Advanced Studies Network Neutrality, Access to Content and Online Advertising Anna D Annunzio and Antonio Russo European University Institute Robert Schuman Centre

More information

ECON 312: Oligopolisitic Competition 1. Industrial Organization Oligopolistic Competition

ECON 312: Oligopolisitic Competition 1. Industrial Organization Oligopolistic Competition ECON 312: Oligopolisitic Competition 1 Industrial Organization Oligopolistic Competition Both the monopoly and the perfectly competitive market structure has in common is that neither has to concern itself

More information

Access Regulation with Asymmetric Termination Costs

Access Regulation with Asymmetric Termination Costs No 29 Access Regulation with Asymmetric Termination Costs Torben Stühmeier July 2011 IMPRINT DICE DISCUSSION PAPER Published by Heinrich Heine Universität Düsseldorf, Department of Economics, Düsseldorf

More information

1.4 Hidden Information and Price Discrimination 1

1.4 Hidden Information and Price Discrimination 1 1.4 Hidden Information and Price Discrimination 1 To be included in: Elmar Wolfstetter. Topics in Microeconomics: Industrial Organization, Auctions, and Incentives. Cambridge University Press, new edition,

More information

INTERCONNECTION AMONG COMPETITORS AND THE REGULATION OF TELECOMMUNICATIONS

INTERCONNECTION AMONG COMPETITORS AND THE REGULATION OF TELECOMMUNICATIONS INTERCONNECTION AMONG COMPETITORS AND THE REGULATION OF TELECOMMUNICATIONS Jean Tirole Institut D Economie Industrielle, University of Toulouse October 10, 2006 Universitat Pompeu Fabra, Facultat de Ciències

More information

Net Neutrality, Boon or Bane for Free File Sharing Service? (preliminary draft)

Net Neutrality, Boon or Bane for Free File Sharing Service? (preliminary draft) Net Neutrality, Boon or Bane for Free File Sharing Service? (preliminary draft) Tong Wang Florence School of Regulation, European University Institute June 29, 2012 Abstract In this paper, we establish

More information

DOC NO: INFOSOC 53/15 DATE ISSUED: June 2015. Resolution on the open and neutral Internet

DOC NO: INFOSOC 53/15 DATE ISSUED: June 2015. Resolution on the open and neutral Internet DOC NO: INFOSOC 53/15 DATE ISSUED: June 2015 Resolution on the open and neutral Internet Introduction This resolution builds on the TACD net neutrality resolutions of April 2010 1 and June 2014 2, which

More information

No 166. Homogeneous Platform Competition with Endogenous Homing. Thomas D. Jeitschko, Mark J. Tremblay

No 166. Homogeneous Platform Competition with Endogenous Homing. Thomas D. Jeitschko, Mark J. Tremblay No 166 Homogeneous Platform Competition with Endogenous Homing Thomas D. Jeitschko, Mark J. Tremblay November 2014 IMPRINT DICE DISCUSSION PAPER Published by düsseldorf university press (dup) on behalf

More information

Pricing and Persuasive Advertising in a Differentiated Market

Pricing and Persuasive Advertising in a Differentiated Market Pricing and Persuasive Advertising in a Differentiated Market Baojun Jiang Olin Business School, Washington University in St. Louis, St. Louis, MO 63130, baojunjiang@wustl.edu Kannan Srinivasan Tepper

More information

Why payment card fees are biased against merchants

Why payment card fees are biased against merchants Why payment card fees are biased against merchants Julian Wright November 2010 Abstract I formalize the popular argument that payment card networks such as MasterCard and Visa charge merchants too much

More information

Working Paper Pricing Internet Traffic: Exclusion, Signalling and Screening

Working Paper Pricing Internet Traffic: Exclusion, Signalling and Screening econstor www.econstor.eu Der Open-Access-Publikationsserver der ZBW Leibniz-Informationszentrum Wirtschaft The Open Access Publication Server of the ZBW Leibniz Information Centre for Economics Jullien,

More information

Technology platforms, such as Microsoft Windows, are the hubs of technology industries. We develop a

Technology platforms, such as Microsoft Windows, are the hubs of technology industries. We develop a MANAGEMENT SCIENCE Vol. 52, No. 7, July 2006, pp. 1057 1071 issn 0025-1909 eissn 1526-5501 06 5207 1057 informs doi 10.1287/mnsc.1060.0549 2006 INFORMS Two-Sided Competition of Proprietary vs. Open Source

More information

Conditions for Efficiency in Package Pricing

Conditions for Efficiency in Package Pricing Conditions for Efficiency in Package Pricing Babu Nahata Department of Economics University of Louisville Louisville, Kentucky 40292, USA. e-mail: nahata@louisville.edu and Serguei Kokovin and Evgeny Zhelobodko

More information

3 Price Discrimination

3 Price Discrimination Joe Chen 26 3 Price Discrimination There is no universally accepted definition for price discrimination (PD). In most cases, you may consider PD as: producers sell two units of the same physical good at

More information

On the Antitrust Economics of the Electronic Books Industry

On the Antitrust Economics of the Electronic Books Industry On the Antitrust Economics of the Electronic Books Industry Germain Gaudin (DICE, Heinrich Heine University Düsseldorf) & Alexander White (School of Economics & Management, Tsinghua University) Postal

More information

Internet access and use

Internet access and use Internet access and use Christian Helmers 21 April 2015 1 / 23 Internet access and use Most digital technologies rely on access to the Internet Who has access? How has access to the Internet changed economic

More information

Pricing in a Competitive Market with a Common Network Resource

Pricing in a Competitive Market with a Common Network Resource Pricing in a Competitive Market with a Common Network Resource Daniel McFadden Department of Economics, University of California, Berkeley April 8, 2002 I. This note is concerned with the economics of

More information

Frequent flyer programs and dynamic contracting with limited commitment

Frequent flyer programs and dynamic contracting with limited commitment Frequent flyer programs and dynamic contracting with limited commitment Emil Temnyalov March 14, 2015 Abstract I present a novel contract theoretic explanation of the profitability and management of loyalty

More information

Remarks on Broadband Competition and Access Regulation

Remarks on Broadband Competition and Access Regulation Remarks on Broadband Competition and Access Regulation Marius Schwartz* FCC Workshop on Economic Issues in Broadband Competition Washington, DC October 9, 2009 * Professor of Economics, Georgetown University,

More information

2. Information Economics

2. Information Economics 2. Information Economics In General Equilibrium Theory all agents had full information regarding any variable of interest (prices, commodities, state of nature, cost function, preferences, etc.) In many

More information

Paul Belleflamme, CORE & LSM, UCL

Paul Belleflamme, CORE & LSM, UCL International Workshop on Supply Chain Models for Shared Resource Management Managing inter- and intra-group externalities on two-sided platforms Paul Belleflamme, CORE & LSM, UCL 22/01/2010 FUSL, Brussels

More information

Cars, Broadband, Internet: And why the road to innovation may go through Washington DC

Cars, Broadband, Internet: And why the road to innovation may go through Washington DC Cars, Broadband, Internet: And why the road to innovation may go through Washington DC Jon M. Peha Carnegie Mellon University www.ece.cmu.edu/~peha/bio.html Keynote for IEEE CCNC / Consumer Electronics

More information

On the optimality of optimal income taxation

On the optimality of optimal income taxation Preprints of the Max Planck Institute for Research on Collective Goods Bonn 2010/14 On the optimality of optimal income taxation Felix Bierbrauer M A X P L A N C K S O C I E T Y Preprints of the Max Planck

More information

Net neutrality and innovation at the core and at the edge

Net neutrality and innovation at the core and at the edge Net neutrality and innovation at the core and at the edge Carlo Reggiani Tommaso Valletti y 15th May 2011 - Very preliminary and incomplete Abstract Net neutrality is a hotly debated topic. A key point

More information

Online Supplementary Material

Online Supplementary Material Online Supplementary Material The Supplementary Material includes 1. An alternative investment goal for fiscal capacity. 2. The relaxation of the monopoly assumption in favor of an oligopoly market. 3.

More information

«Internet access and investment incentives for broadband service providers»

«Internet access and investment incentives for broadband service providers» «Internet access and investment incentives for broadband service providers» Edmond BARANES Jean-Christophe POUDOU DR n 2011-09 Internet access and investment incentives for broadband service providers

More information

The economics of network neutrality

The economics of network neutrality RAND Journal of Economics Vol. 43, No. 4, Winter 2012 pp. 602 629 The economics of network neutrality Nicholas Economides and Benjamin E. Hermalin Under the current regime for Internet access, network

More information

Oligopoly: Cournot/Bertrand/Stackelberg

Oligopoly: Cournot/Bertrand/Stackelberg Outline Alternative Market Models Wirtschaftswissenschaften Humboldt Universität zu Berlin March 5, 2006 Outline 1 Introduction Introduction Alternative Market Models 2 Game, Reaction Functions, Solution

More information

Please Do not Circulate Preliminary and Incomplete Mixed Bundling and Imperfect Competition

Please Do not Circulate Preliminary and Incomplete Mixed Bundling and Imperfect Competition Abstract A model of two-product, Hotelling duopolists is examined in which firms can engage in mixed bundling. Even though products are independent in consumption, bundled pricing induces complementarity

More information

INDIAN INSTITUTE OF MANAGEMENT CALCUTTA WORKING PAPER SERIES. WPS No. 681/ September 2011

INDIAN INSTITUTE OF MANAGEMENT CALCUTTA WORKING PAPER SERIES. WPS No. 681/ September 2011 INDIAN INSTITUTE OF MANAGEMENT CALCUTTA WORKING PAPER SERIES WPS No. 681/ September 2011 Pricing Infrastructure-as-a-Service for Online Two- Sided Platform Providers by Soumyakanti Chakraborty Assistant

More information

Chapter 11 Pricing Strategies for Firms with Market Power

Chapter 11 Pricing Strategies for Firms with Market Power Managerial Economics & Business Strategy Chapter 11 Pricing Strategies for Firms with Market Power McGraw-Hill/Irwin Copyright 2010 by the McGraw-Hill Companies, Inc. All rights reserved. Overview I. Basic

More information

Pricing Payment Cards

Pricing Payment Cards Pricing Payment Cards Özlem Bedre Emilio Calvano November 2008 Abstract In a payment card association such as Visa, each time a consumer pays by card, the bank of the merchant (acquirer) pays an interchange

More information

Regulation is used to foster and sustain competition

Regulation is used to foster and sustain competition General Regulation Regulation is used to foster and sustain competition Competitiveness Public Monopoly Need for regulation Limited regulation: Government supplied telecoms services Private Monopoly Increased

More information

November 5, 2014 BY ECFS. Ms. Marlene Dortch Federal Communications Commission 445 Twelfth Street, SW Washington, DC 20554

November 5, 2014 BY ECFS. Ms. Marlene Dortch Federal Communications Commission 445 Twelfth Street, SW Washington, DC 20554 November 5, 2014 BY ECFS Ms. Marlene Dortch Federal Communications Commission 445 Twelfth Street, SW Washington, DC 20554 Re: Notice of Ex Parte Submission, Protecting and Promoting the Open Internet,

More information

Subsidies in oligopoly markets: a welfare comparison between symmetric and asymmetric costs

Subsidies in oligopoly markets: a welfare comparison between symmetric and asymmetric costs Subsidies in oligopoly markets: a welfare comparison between symmetric and asymmetric costs Stephen F. Hamilton and Rickard Sandin Oligopolistic industries generally produce less than is socially desirable.

More information

Network Neutrality and Competition on Advertising Markets -Preliminary-

Network Neutrality and Competition on Advertising Markets -Preliminary- Network Neutrality and Competition on Advertising Markets -Preliminary- Anna D'Annunzio Antonio Russo 0 March 013 Abstract We investigate the eects of removing network neutrality regulation on the distribution

More information

-1- Worked Solutions 5. Lectures 9 and 10. Question Lecture 1. L9 2. L9 3. L9 4. L9 5. L9 6. L9 7. L9 8. L9 9. L9 10. L9 11. L9 12.

-1- Worked Solutions 5. Lectures 9 and 10. Question Lecture 1. L9 2. L9 3. L9 4. L9 5. L9 6. L9 7. L9 8. L9 9. L9 10. L9 11. L9 12. -1- Worked Solutions 5 Lectures 9 and 10. Question Lecture 1. L9 2. L9 3. L9 4. L9 5. L9 6. L9 7. L9 8. L9 9. L9 10. L9 11. L9 12. L10 Unit 5 solutions Exercise 1 There may be practical difficulties in

More information

Working Paper Series

Working Paper Series RGEA Universidade de Vigo http://webs.uvigo.es/rgea Working Paper Series A Market Game Approach to Differential Information Economies Guadalupe Fugarolas, Carlos Hervés-Beloso, Emma Moreno- García and

More information

Optimal Age Specific Income Taxation

Optimal Age Specific Income Taxation Optimal Age Specific Income Taxation Jean-Marie Lozachmeur 1 October, 2005 1 GREMAQ, University of Toulouse. Mailing: Gremaq, 21 Allée de Brienne F- 31000 Toulouse. E-mail: jean-marie.lozachmeur@univ-tlse1.fr.

More information

Economics Discussion Paper Series EDP-1202

Economics Discussion Paper Series EDP-1202 Economics Discussion Paper Series EDP-1202 Net neutrality and innovation at the core and at the edge Carlo Reggiani Tommaso Valletti January 2012 Economics School of Social Sciences The University of Manchester

More information

Performance of networks containing both MaxNet and SumNet links

Performance of networks containing both MaxNet and SumNet links Performance of networks containing both MaxNet and SumNet links Lachlan L. H. Andrew and Bartek P. Wydrowski Abstract Both MaxNet and SumNet are distributed congestion control architectures suitable for

More information

Bilateral monopoly in telecommunications: bargaining over fixed-to-mobile termination rates

Bilateral monopoly in telecommunications: bargaining over fixed-to-mobile termination rates Bilateral monopoly in telecommunications: bargaining over fixed-to-mobile termination rates Tommaso Majer October 2009 Departament d Economia i d Història Econòmica, Edifici B Universitat Autònoma de Barcelona

More information

Bundling in Cable Television: A Pedagogical Note With a Policy Option. Keith Brown and Peter J. Alexander Federal Communications Commission, USA

Bundling in Cable Television: A Pedagogical Note With a Policy Option. Keith Brown and Peter J. Alexander Federal Communications Commission, USA K. Bundling Brown and in Cable P. J. Alexander Television Bundling in Cable Television: A Pedagogical Note With a Policy Option Keith Brown and Peter J. Alexander Federal Communications Commission, USA

More information

Issue Brief: How to Deal with Data Caps, Sponsored Data and Zero- Rating

Issue Brief: How to Deal with Data Caps, Sponsored Data and Zero- Rating Issue Brief: How to Deal with Data Caps, Sponsored Data and Zero- Rating Matt Wood Free Press The FCC adopted its landmark Open Internet Order nearly a year ago. In this the agency made the correct decision

More information

Regulation of fixed and mobile termination charges

Regulation of fixed and mobile termination charges Regulation of fixed and mobile termination charges Amsterdam, 27 February 2013 Commissioned by UPC Nederland and Ziggo Regulation of fixed and mobile termination charges A comparison of pure BULRIC and

More information

Work incentives and household insurance: Sequential contracting with altruistic individuals and moral hazard

Work incentives and household insurance: Sequential contracting with altruistic individuals and moral hazard Work incentives and household insurance: Sequential contracting with altruistic individuals and moral hazard Cécile Aubert Abstract Two agents sequentially contracts with different principals under moral

More information

Regret and Rejoicing Effects on Mixed Insurance *

Regret and Rejoicing Effects on Mixed Insurance * Regret and Rejoicing Effects on Mixed Insurance * Yoichiro Fujii, Osaka Sangyo University Mahito Okura, Doshisha Women s College of Liberal Arts Yusuke Osaki, Osaka Sangyo University + Abstract This papers

More information

Pros and cons of net neutrality: an economic assessment

Pros and cons of net neutrality: an economic assessment Pros and cons of net neutrality: an economic assessment Dr. Hans W. Friederiszick Dr. Simone Kohnz DT Representation Berlin, 18 March 2011 This presentation is based on a study by Friederiszick/ Grajek/

More information

Strategic Differentiation by Business Models: Free-to-air and Pay-TV s

Strategic Differentiation by Business Models: Free-to-air and Pay-TV s Strategic Differentiation by Business Models: Free-to-air and Pay-TV s Emilio Calvano (CSEF - U. of Naples) and Michele Polo (U. Bocconi, IEFE and IGIER) October 2014 - Naples. 12th Conference on Media

More information

Why payment card fees are biased against retailers

Why payment card fees are biased against retailers Why payment card fees are biased against retailers Julian Wright June 2012 Abstract I formalize the popular argument that retailers pay too much and cardholders too little to make use of payment card platforms,

More information

Economics of Insurance

Economics of Insurance Economics of Insurance In this last lecture, we cover most topics of Economics of Information within a single application. Through this, you will see how the differential informational assumptions allow

More information

Net Neutrality and Investment Incentives

Net Neutrality and Investment Incentives Net Neutrality and Investment Incentives Jay Pil Choi y Byung-Cheol Kim z March 9, 00 Abstract This paper analyzes the e ects of net neutrality regulation on investment incentives for Internet service

More information

Intellectual Property Right Protection in the Software Market

Intellectual Property Right Protection in the Software Market Intellectual Property Right Protection in the Software Market Yasuhiro Arai Aomori Public College 153-4, Yamazaki, Goushizawa, Aomori-city, Aomori 030-0196, Japan March 01 Abstract We discuss the software

More information

Submission by the Asia Pacific Carriers Coalition

Submission by the Asia Pacific Carriers Coalition Submission by the Asia Pacific Carriers Coalition In Response to Consultation Paper issued by TRAI on Relaxing Restrictive Provision of Internet Telephony (IPT) (Consultation Paper No. 11/08 issued on

More information

On the Interaction and Competition among Internet Service Providers

On the Interaction and Competition among Internet Service Providers On the Interaction and Competition among Internet Service Providers Sam C.M. Lee John C.S. Lui + Abstract The current Internet architecture comprises of different privately owned Internet service providers

More information

ANOTHER PERVERSE EFFECT OF MONOPOLY POWER

ANOTHER PERVERSE EFFECT OF MONOPOLY POWER ANOTHER PERVERSE EFFECT OF MONOPOLY POWER Jean J. Gabszewicz and Xavier Y. Wauthy November 2000 Abstract We show that the simple fact that a monopolist sells a good in units which are indivisible may well

More information

A Model-based Study of the Impact on the Internet of Managed Services and the Spawning of User Innovated Applications

A Model-based Study of the Impact on the Internet of Managed Services and the Spawning of User Innovated Applications A Model-based Study of the Impact on the Internet of Managed Services and the Spawning of User Innovated Applications Debasis Mitra Department of Electric Engineering Columbia University debasismitra@columbia.edu

More information

Credible Discovery, Settlement, and Negative Expected Value Suits

Credible Discovery, Settlement, and Negative Expected Value Suits Credible iscovery, Settlement, and Negative Expected Value Suits Warren F. Schwartz Abraham L. Wickelgren Abstract: This paper introduces the option to conduct discovery into a model of settlement bargaining

More information

Optimal health insurance contract: Is a deductible useful?

Optimal health insurance contract: Is a deductible useful? Economics Letters 87 (2005) 313 317 www.elsevier.com/locate/econbase Optimal health insurance contract: Is a deductible useful? David Bardey a, T, Romain Lesur b a GREMAQ UMR CNRS 5604-Manufacture des

More information

Net Neutrality & VoIP Carriers ECG.CO, NOT A LAW FIRM

Net Neutrality & VoIP Carriers ECG.CO, NOT A LAW FIRM Net Neutrality & VoIP Carriers MARK R LINDSEY, NOT A LAWYER ECG.CO, NOT A LAW FIRM Background 2005: ISP AT&T announces preference for sender pays, & special fees for more for preferential treatment of

More information

Oligopoly: How do firms behave when there are only a few competitors? These firms produce all or most of their industry s output.

Oligopoly: How do firms behave when there are only a few competitors? These firms produce all or most of their industry s output. Topic 8 Chapter 13 Oligopoly and Monopolistic Competition Econ 203 Topic 8 page 1 Oligopoly: How do firms behave when there are only a few competitors? These firms produce all or most of their industry

More information

Optimal election of qualities in the presence of externalities

Optimal election of qualities in the presence of externalities Optimal election of qualities in the presence of externalities Dolores Garcia Maria Tugores October 2002 Abstract In this paper we obtain the quality and the level of production that a certain sector would

More information

Network Neutrality Revisited: Challenges and responses in the EU and in the US

Network Neutrality Revisited: Challenges and responses in the EU and in the US Network Neutrality Revisited: Challenges and responses in the EU and in the US J. Scott Marcus The opinions expressed are the sole responsibility of the author and do not necessarily represent the official

More information

Interconnection in Network Industries

Interconnection in Network Industries Review of Industrial Organization 14: 1 25, 1999. 1999 Kluwer Academic Publishers. Printed in the Netherlands. 1 Interconnection in Network Industries MICHAEL CARTER and JULIAN WRIGHT Department of Economics,

More information

Is it too soon to declare cable the winner in the broadband wars?

Is it too soon to declare cable the winner in the broadband wars? Is it too soon to declare cable the winner in the broadband wars? New technologies give telecoms a boost in their battle against cable to deliver high-speed broadband to America s homes with big implications

More information

Review of Card Surcharging: A Consultation Document

Review of Card Surcharging: A Consultation Document Review of Card Surcharging: A Consultation Document june 2011 Contents 1. Introduction 1 2. Background 2 3. Concerns about Surcharging Practices 5 4. Policy Options and Discussion 7 5. Summary of Issues

More information

Sharing Online Advertising Revenue with Consumers

Sharing Online Advertising Revenue with Consumers Sharing Online Advertising Revenue with Consumers Yiling Chen 2,, Arpita Ghosh 1, Preston McAfee 1, and David Pennock 1 1 Yahoo! Research. Email: arpita, mcafee, pennockd@yahoo-inc.com 2 Harvard University.

More information

Lobbying on Entry Regulations under Imperfect. Competition

Lobbying on Entry Regulations under Imperfect. Competition Lobbying on Entry Regulations under Imperfect Competition Dapeng CAI a and Shinji KOBAYASHI b a Corresponding Author. Institute for Advanced Research, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-86,

More information

Credit card interchange fees

Credit card interchange fees Credit card interchange fees Jean-Charles Rochet Julian Wright January 23, 2009 Abstract We build a model of credit card pricing that explicitly takes into account credit functionality. We show that a

More information

* University of Notre Dame, {hguo, arthurlim}@nd.edu. ** University of Florida, {shubho.bandyopadhyay, kenny.cheng}@warrington.ufl.

* University of Notre Dame, {hguo, arthurlim}@nd.edu. ** University of Florida, {shubho.bandyopadhyay, kenny.cheng}@warrington.ufl. EFFECTS OF COMPETITION AMONG INTERNET SERVICE PROVIDERS AND CONTENT PROVIDERS ON THE NET NEUTRALITY DEBATE Hong Guo *, Subhajyoti Bandyopadhyay **, Arthur Lim *, Yu-Chen Yang ***, Kenny Cheng ** * University

More information

Quality of Service (QoS)

Quality of Service (QoS) Net Neutrality Scott Jordan Department of Computer Science University of California, Irvine Quality of Service (QoS) more interactive less interactive telephone calls video conferencing Internet radio

More information

9.1 Cournot and Bertrand Models with Homogeneous Products

9.1 Cournot and Bertrand Models with Homogeneous Products 1 Chapter 9 Quantity vs. Price Competition in Static Oligopoly Models We have seen how price and output are determined in perfectly competitive and monopoly markets. Most markets are oligopolistic, however,

More information